The ranking in this dataset covers 100 companies, with a single leader holding 778 records well ahead of a fifth-place figure of 304 and a tenth-place figure of 123. That leader-to-tail gap means one company has built a materially deeper filing position than the rest of the ranked group. The top five assignees combined hold 35.0% of all 6,722 records in scope, so a small number of firms account for a disproportionate share of the field even though the ranking runs one hundred names deep.

Filings grew from 270 in 2017 to a peak of 1,072 in 2023, then eased to 779 in 2024 — a 16% drop across that three-year span from 2021's 924. That is the most recent period that can be treated as complete, because publication lags filing by roughly 18 months and the 2025-2026 figures are still filling in. Read the recent pullback as a real move off the 2023 peak, not as evidence the field is dying out, since later years will revise upward as more filings publish.

The dominant IPC class is G06N, computing based on AI models, appearing on 87.5% of all 6,722 records, followed by G06F electric digital data processing at 27.3%. Narrower classes tied to the physical execution layer — B82Y nanotechnology, H01L semiconductors, H10N solid-state devices, H03K pulse technique, and H04L digital transmission — each sit at under 10% of records. Because a single record can carry several IPC codes, these shares add up to well over 100%, which is expected and not a data error.

The class data suggests thinner filing activity around the physical control layer relative to the algorithmic core: H03K pulse-technique circuits sit at 5.6% of records, H10N solid-state devices at 5.9%, and H04L digital-transmission interfaces at 5.2%, despite the search scope explicitly including control-pulse and qubit-control terms. That gap points toward multi-qubit control-pulse calibration, cross-modal coupling architectures, and quantum-output transmission interfaces as areas with comparatively less claim density than the core algorithmic and AI-model classes. A first claim in these areas would need to tie a specific pulse-calibration or transmission mechanism to a defined circuit or coupling structure to differentiate from the dense G06N core.

Citation counts here favour older records simply because they have had more time in circulation to accumulate citations within the searched corpus, so a high count signals influence on later filings rather than current commercial relevance. The most-cited record in this dataset, US20220366494A1, deals with market orchestration rather than core quantum hardware, which illustrates why citation rank and technical centrality do not always align. Use citation counts as one input alongside filing recency and assignee activity, not as a standalone importance score.